Assessing predictive performance of models in the COVID-19 Forcast Hub
Estee Y Cramer, Nicholas G Reich,
2020-10-15
Prelims
Figure 2: Model Performance stratified by target (aggregated by week and location): emperical and calculated values
Figure 3: average performance of each model by location (emperical estimates)
Figure 4: average performance of each model by week (averaged across locations, not imputed for weeks that are missing)
Figure 5: Average performance of incident models by target
Figure 6: average performance of each model by location (Incident Deaths)
Figure 7 Average performance stratified by week of submission